Skip to main content

Full text of "Observations on Gigantiops destructor Fabricius and other leaping ants."

See other formats

A; 21 



[Reprinted from BIOLOGICAL BULLETIN, Vol. XLII., No, 4, April, 1922.) 

[Reprinted from BIOLOGICAL BULLETIN, Vol. XLII., No. 4, April, 1922.1 



In any study of the very exuberant ant-fauna of the Neotropical 
Region one can not fail to be impressed by the striking contrast 
between certain genera like Eciton, Pseudomyrma, Solenopsis, 
Crematogaster, Cryptocercus, Azteca, and Camponotus, each rep- 
resented by a large number of variable species, and genera like 
Paraponera, Acanthognathus, Daceton, Blepharidatta, Stegomyr- 
mex, and Gigantiops, each represented by a single, very stable 
species. Of course, such monotypic groups may be regarded 
either as very ancient, embracing during some former age many 
species of which only one has survived, or as single species which, 
after acquiring generic status in the remote past, have since under- 
gone little or no modification. The individuals of a species repre- 
senting a monotypic genus may be either very rare or local, mere 
relicts of a bygone age, or prominent and ubiquitous over larger 
geographical areas. This is true of such ants as Paraponera 
clavata Fabr. and Gigantiops destructor, which I have recently had 
abundant opportunity to study in the jungle about the Tropical 
Laboratory of the New York Zoological Society at Kartabo, Brit- 
ish Guiana. As the latter species is the more imperfectly known, 
I have singled it out for special consideration. 

The name Gigantiops destructor conjures up visions of a huge- 
eyed, insatiable monster, a kind of Cyclopean insect-jaguar. Fa- 
bricius, when he first described the insect in 1804 as Formica 
destructor, certainly knew nothing of its behavior and probably 
gave it what seemed to him an appropriate specific name for any 
ant measuring a centimeter in length. More than half a century 
later ('58) Frederick Smith received specimens taken by Bates at 
Ega, Brazil, and believed them to represent a new species which he 
described as Formica solitaria. The following note was appended 

i Contributions from the Entomological Laboratory of the Bussey Institu- 
tion, Harvard University. No. 177. 



to the description: "This is a very remarkable insect; for inde- 
pendent of the enormously developed eyes and produced clypeus, 
the palpi are elongated to half the length of the thorax, the max- 
illary are six-, and the labial four-jointed. Mr. Bates says : 
"This curious solitary ant is never seen by more than one at a 
time, prowling about fallen leaves, etc., in the forest ; I have never 
seen its formicarium and, from its solitary habits, have no clue to 
guide me in looking for it." Perhaps Smith was confirmed in his 
choice of the specific name by the monastic or ascetic appearance 
of the insect, its somber black livery, relieved only by the golden- 
yellow tips of its antennae, its long, emaciated limbs and its huge 
eyes, perpetually dilated as if in astonishment and chagrin at the 
indecent behavior of other insects. 

Two short notes, however, by later observers indicate that 
Gigantiops may be neither an insatiable assassin nor a humble 
anchorite, but a harmless and perhaps rather frivolous creature, 
that may have become permanently goggle-eyed through an age- 
long endeavor to enjoy to the full the riotous beauties of its en- 
vironment. Emery ('93) was informed by Albert Schulz that the 
" Brazilian ant, Gigantiops destructor Fabr., which is distinguished 
by its enormous eyes, leaps from twig to twig, like the Odonto- 
machus hcematodes living in the same places," and Mann (1916) 
says : " In life this is one of the most attractive ants encountered. 
It lives always in the forest, where it forages either among the 
branches of trees or on ^he ground. The movements of the forag- 
ing worker are rapid, comparable to some of our species of Cicin- 
dela, and the bicolored antennae are kept constantly in motion." 

Roger ('63) was the first to throw Smith's Formica solitaria 
into the synonymy and to establish the peculiar genus Gigantiops. 
In more recent myrmecological literature mention of the insect 
recurs sporadically and at long intervals, showing that it was rarely 
seen in the many collections of South American ants examined by 
Mayr, Forel, Emery, Santschi, and others. Its known range, as 
indicated by the literature and by specimens in my collection, is 
as follows : 

Brazil: Ega (Bates) ; Para (E. Goeldi, ex coll. Forel) ; Maran- 
hao (Ducke) ; Para, Abuna, Porto Velho, and Madeira-Mamore 
R. R. (W. M. Mann).. 


French Guiana (Jelski). 

British Guiana: Kaieteur Falls, Tukheit, and Tumatumari .(F. 
E. Lutz) ; Penal Settlement, Bartica District (W. Beebe) ; Kar- 
tabo and Kalacoon (Wheeler) . 

Peru: Callanga (S'taudinger). 

Bolivia: Rio Beni (L. Balzan). 

These localities show that Gigantiops has. a very limited range 
compared with many Neotropical ants, since it is confined to a 
strip of South America east of the Andes and extending from 
about 10 north to 10 south of the equator. 1 

Gigantiops (Fig. i) is a common ant in the forested portions of 
British Guiana, preferring 'shady places rather free from under- 
growth and spending most of its time on the ground, running over 
the dead leaves. It occurs singly, as stated by previous observers, 
and really belongs to a forest-floor ant-fauna comprising also 
Neoponera apicalis Latr., obscuricornis Emery and commutata 
Roger, Mesoponera constricta Mayr, Pachycondyla crassinoda La- 
treille and harp ax Fabr., Paraponera clavata Fabr., and Ectatomma 
quadridens Fabr. Those who are interested in mimicry will ob- 
serve that in its form, the dull black color of its body and yellow 
antennal tips, Gigantiops bears such a striking resemblance to 
N. apicalis and obscuricornis that the latter might be regarded as 

1 1 find a note by von Motschulsky in a letter published in his " Etudes 
Entomologiques " (1855) and referred to in the "Stettiner Entomologische 
Zeitung " (1859), which seems to apply to Gigantiops, Speaking of the insects 
which he observed at Obispo, Panama, he says : " I observed a lot of ants of 
diverse and bizarre form, among others one bearing the closest resemblance to 
a spider, especially to a Salticus, and as it also has the ability to leap, I have 
named it Salticomorpha nigra." There are two objections to accepting the 
name Salticomorpha as antedating Gigantiops: first, there is no record of this 
insect's having been taken in Central America or even in Colombia, and 
second, von Motschulsky may have seen a Pseudomyrma gracilis Fabr., which 
resembles a black Attid spider in form and color, and have mistaken its 
erratic movements for leaps. The well-known arachnologist, E. Simon (in 
Emery, "Voyage de M. E. Simon" (Dec., i887-Avril, 1888). Formicides, 
Ann. Soc. Ent. France 1890, p. 65 nota) noticed that " all the species of the 
genus Pseudomyrma reproduce exactly the forms and colors of the spiders of 
the genus Simonella Peckh. (Attidae) and the resemblance is equally striking 
in their gait." For the present it seems advisable, therefore, either to treat 
Salticomorpha nigra Motsch, as a nomen midum or to include it with a query 
in the synonymy of Gigantiops destructor Fabr. 



its models. Furthermore, these Ponerines sting very severely, 
whereas Gigantiops can be picked up with impunity. In Kartabo, 
nevertheless, the models are much less frequently seen than the 
mimic. This is interesting in connection with the observations of 
Mr. Tee Van, who finds that in the same region many of the 

FIG. i. Gigantiops destructor Fabr. Worker, about twice natural size ; 
dorsal and lateral views and head from above. 

mimetic butterflies are much more abundant than their putative 
Heliconid models. It would be a mistake to suppose that Gi- 
gantiops acts as if it derived any benefit from its striking resem- 
blance to the stinging Ponerines. It greatly surpasses them in 
agility and when pursued will even leap several inches in a very 
graceful, cat-like manner. On the rather infrequent occasions 
when it climbs onto bushes and is running over their foliage it will, 
if disturbed, leap, without the slightest hesitation, to another leaf 
or even to the ground and make off with great alacrity. When two 
Gigantiops happen to meet face to face, they exhibit a peculiar 
play. After stroking each other's heads for a moment with the 
yellow tips of their antennae, they move from side to side, precisely 
like two persons who meet on the sidewalk and try to prevent each 
other from passing. 


On rare occasions Gigantiops may be seen carrying a termite 
worker or other small insect in its jaws, but even such individuals 
are not easily followed to their nests. Forel's placing of this ant 
in his tribe (Ecophyllini, i.e., with (Ecophylla smaragdina Fabr., 
the well-known tree-ant of the Old World tropics, naturally led 
me to suppose that the nest must be in the trees, but this supposi- 
tion, which has probably been shared by other myrmecologists, 
proves to be erroneous. On July 14, after much careful search 
and persistent following of single workers, my son Ralph suc- 
ceeded in finding a nest in a partly decayed log only three or four 
inches in diameter lying on the ground at the edge of the Puruni 
trail and brought the portion inhabited by the ants into the labo- 
ratory. As soon as I began to dig into their nest the workers 
leaped out and made off, holding their larvae in their mandibles. 
The colony comprised only fifty or sixty workers, which had been 
living in some large cavities made by Passalus or other wood- 
inhabiting beetles. In one of the chambers there were empty 
cocoons, showing that the pupae of Gigantiops are not nude as in 
(Ecophylla, its supposed nearest ally among the Formicinae. I 
failed to secure the queen and believe she must have escaped un- 
observed among the workers. From these she differs merely in 
her somewhat larger size and slightly more voluminous thorax. 

Notwithstanding careful search by my son and myself, ten days 
elapsed before I could again observe a Gigantiops nest. I found 
the second nest in a similar situation, in a partly decayed piece of 
a Cecropia trunk about a foot and a half long and three inches in 
diameter, lying loosely on the dead leaves in the shade of a bush. 
I noticed one of the workers timidly guarding a small hole and 
hastily retreating into it on my approach. The hole was plugged 
with cotton and the log carried back to the laboratory. In order 
that the ants might not elude me as on the previous occasion, I 
opened the log over a pail of water, but notwithstanding these 
precautions a few of the workers managed to escape. The entire 
colony, which was inhabiting one of the large internodal cavities 
so peculiar to Cecropia, was scarcely larger than the former colony, 
but contained more larvae and several freshly spun worker cocoons. 

1 90 


No males were present and future attempts to find them and other 
colonies were unsuccessful. 1 

Two matters call for further discussion in connection with the 
foregoing observations, the taxonomic affinities of Gigantiops de- 
structor and its modus saltandi. The generally accepted view of 
its taxonomic position can be traced to the various papers which 
Forel has published from time to time on the classification of the 
subfamily Formicinse (= Camponotinse Forel) . In 1878 he placed 
the genus Gigantiops between Opisthopsis and CEcophylla in his 
first tribe of the subfamily. In his classification of 1893 he 
omitted all mention of Gigantiops, though he enumerated the vari- 
ous other genera of the subfamily. In 1912 he remodeled .the 
classification and considerably augmented the number of tribes, to 
one of which, the CEcophyllini, he assigned the three genera 
Gigantiops, Myrmecorhynchus, and CEcophylla. The same arrange- 
ment is preserved in his paper of 1917. I endeavored to show in 
the same year that Myrmecorhynchus could not be retained among 
the CEcophyllini, but should probably constitute an independent 
tribe, the Myrmecorhynchini. The characters of the CEcophyllini, 
according to Forel, are the following: gizzard long and narrow, 
with straight calyx; clypeal fossa mo're or less distinct from the 
antennary fossa; antennas inserted a little behind the frontal area, 
but near the anterior ends of the frontal carinae. The gizzard 
characters are not peculiar to this tribe, but recur also in the 
Camponotini, and the remaining characters are decidedly weak, 
since they depend on slight differences in the proportions of the 
anterior portions of the head. When we compare Gigantiops with 
CEcophylla we are struck by the great differences in the structure 
of the larva, pupa and adult and in habits. That the habits of the 
two ants are totally different will be seen from a comparison of 
the observations above recorded with what we know of CEcophylla, 
and its various subspecies and varieties, which are arboreal ants 
inhabiting peculiar nests made of leaves and silk spun by their 
larvae. Still it may be objected that such ethological peculiarities 
have little significance, since we have species of Camponotus that 

i Dr. W. M. Mann, who has just returned with the Mulford Expedition 
from Bolivia, informs me that he found Gigantiops nesting under stones in 
the forests of the Rio Beni. 


live in similar nests (C. senex Smith and formiciformis Forel) 
and others that live in the ground or in rotten logs (C. maculatus 
Fabr. and herculeanus DeGeer) . Turning to morphological char- 
acters, which the taxonomist regards as much more reliable, we 
find that the only resemblances between Gigantiops and CEcophylla 
(apart from the shape of the gizzard which both share with Cam- 
ponotus) are the shape of the clypeus, with its great, projecting 
lobe, the shape of the mandibles, and the feeble characters cited'by 
Forel* There are great differences in the size of the eyes and 
claws, in the shape of the thorax and petiole of the worker, and in 
the size and shape of the thorax of the female, though. the venation 
of the wings is similar. The larva of Gigantiops is like that of 
Camponotus, but very different from that of CEcophylla, and the 
pupa is inclosed in a cocoon. Probably the male Gigantiops will 
be found to exhibit some peculiar differences. Emery (in lift.} 
calls my attention to the singular fact that the tarsal claws of the 
male CEcophylla are almost completely atrophied. It would there- 
fore be very interesting to know the condition of these organs in 
the corresponding sex of Gigantiops. The foregoing considera- 
tions seem to me to render it advisable to remove Gigantiops from 
Forel's tribe (Ecophyllini and to provide an independent tribe for 
its accommodation. I find that Ashmead in 1905 had created such 
a tribe " Gigantiopini," though he included it in a subfamily Geso- 
myrmicinse, with Gesomyrmex and Myrmoteras, genera which, in 
my opinion, are only remotely related to Gigantiops. 

It is practically certain that Gigantiops is one of a number of 
ancient, large-eyed, active Formicinse, once of very wide distribu- 
tion, but now narrowly confined to the tropics. This group, which 
embraces also the genera CEcophylla, Dimorphomyrmex, Geso- 
myrmex, Opisthopsis, Santschiella, and Myrmoteras, represents 
merely the surviving specialized tips of diverging branches of a 
primitive stock. In regard to CEcophylla, Gesomyrmex, arid Di- 
morphomyrmex, we are actually in possession of considerable 
paleontological information. Mayr ('68), Emery ('05), and I 
('14) have recorded the occurrence of two species of each of these 
three genera in the Baltic amber, of Lower Oligocene age ; Emery 
('91) has recorded an CEcophylla and a species allied to Gesomyr- 
mex (Sicelomyrmex Wheeler) from the Sicilian amber, which is 


referred to the Middle Miocene; Forster ('91) a species of CEco- 
phylla from the Middle Oligocene of Alsace, and Heer ('49) and 
Mayr ('67) two species from the Lower Miocene of Croatia. 1 
More recently Cockerell ('20) described a species of this genus 
from the Eocene of England, and he ('15) and Donisthorpe 
('20) three species from the Middle Oligocene of the same 
country. The one, or possibly two, extant species of (Ecophylla 
are now confined to the hottest portions of the Ethiopian, Indo- 
malayan, and Papuan Regions. Similarly the few extant species 
of Gesomyrmex and Dimorphomynnex are known to occur only 
in Borneo and the Philippines. 2 In the same regions and in 
Burma we find the four species of Myrmoteras. Santschiella is 
known only from a single specimen taken in the Belgian Congo, 
the species of the genus Opi'sthopsis are confined to the Aus- 
tralian and Papuan Regions, and the Neotropical Region possesses 
only one of these ancient large-eyed Formicines, Gigantiops. 
This, as we have seen, has a rather limited range and is in all prob- 
ability a true tropical relict, originally developed in and since 
mainly confined to that portion of the ancient South American con- 
tinent known as Archiguiana. All of the genera above mentioned 
are forest ants and most of them are arboreal, but, as we have 
seen, Gigantiops spends most of its time on the forest floor and 
nests in small, partly decayed logs. Opisthopsis nests under bark, 
in the ground or in earthen termitaria, and I may add that Dr. 
F. X. Williams, who took the types of Myrmoteras williamsi 
Wheeler in the Philippines, informs me that this ant nests in the 
soil. From what we know, therefore, of the living and extinct 
forms, we are justified in concluding that the large-eyed Formicinse 
originated during the early Tertiary, or more probably during the 
Cretaceous, and that the extant forms have since undergone little 
or no modification, owing to the very stable ecological conditions 
in which they were able to survive. Gigantiops, in particular, may 

* Since the completion of this paper Professor Cockerell ('21) has described 
a peculiar large-eyed ant from the Green River Eocene of Wyoming as Eoform- 
ica eocenica. It seems to belong to the subfamily Fomicinse and resembles the 
Australian Opisthopsis in the shape of the head and the position of the promi- 
nent eyes. 

2 Since this paper was written I have published the description of a 
Gesomyrmex (G. howardi) from China. 


be said to have an even more remote origin than the archaic though 
highly specialized Neotropical vertebrates, such as the opossums, 
manatees, sloths, armadillos, ant-eaters and tapirs among mammals, 
or the ostriches and hoatzins among birds. 

The jumping or leaping habits of Gigantiops are so unusual that 
a more general account of this behavior as it occurs in various 
Formicidse may not be out of place. There are two very different 
kinds of leaping ants, one which I shall call " retrosalient," which 
always leaps backward, and one that may be called " prosalient," 
because it always leaps forward. Some authors regard the former 
as not " leaping," in the proper sense of the term, probably because 
of the direction and because it is not performed by means of the 
legs. But such very abrupt displacements of the body afe'effected 
in so many different ways in different insects, as, e.g., in Elaterid 
beetles, Lepismids, Collembolans, cheese-maggots, fruit fly and 
VermUeo larvee, the extraordinary Coleopteran (?) cocoons de- 
scribed by Berlese ('20, p. 631), etc., that such words as "leaping" 
can hardly be avoided without pedantry. We even speak of fish 
or of a cataract "leaping." 

Retrosalience has been repeatedly observed in two quite unre- 
lated groups of ants, one embracing the Ponerine genera Odonto- 
machus and Anochetus, the other the Myrmicine genus Stru- 
migenys. By convergence both of these groups have developed 
very similar long, straight, and linear mandibles, inserted close 
together on the front of the head and furnished with large, abruptly 
incurved teeth at their tips. When excited these ants open their 
mandibles so widely that they stand out at right angles to the long 
axis of the head or are even directed slightly backwards. And if 
one of the insects comes in contact with a solid object in its path, 
it closes them so suddenly and with such force that they make an 
audible "click" and the insect is thrown backwards through the 
air to a distance of several inches. I have described this behavior 
in detail in 0. clarus of Texas ('oo). It has also been observed 
in 0. chelifer of Brazil by Schupp (Wasmann, '92) and in the 
common tropicopolitan 0. hamatoda by Nietner ('58), Ferguson 
(Wroughton, '92), Forel, myself, and others. 1 The method of 

i Borgmeier ('20) has recently described the similar habits of 0. affinis in 
Brazil. " They strike their mandibles against the solid substratum and at 
the same moment leap 30 to 35 cm. vertically into the air." 


leaping in Anochetus is precisely similar, as shown by the observa- 
tions of Wroughton ('92) on A. sedillotii var. indicus of India, 
and of Biro (Emery, '97) on a Papuan species. Among the Dace- 
tonini it has been observed by Hetschko in the Brazilian Stru- 
migenys saliens (Mayr, '93) and by 'Biro ('97) in 5". chyzeri 
Emery ('97) of New Guinea. The worker of the latter species 
is able to leap backward to a distance of 20-25 cm., or about 100 
to 150 times the length of its body, but this behavior is not ex- 
hibited by the female. Forel ('93) believed that the Neotropical 
Acanthognathus ocellatus Mayr and Daceton armigerum Latr. 
might be able to leap in the same manner. I have failed to observe 
the habit in the allied Australian ants of the genus Orectognathus. 
All the retrosalient ants, however, leap rather reluctantly and only 
under certain conditions, and the length of their leaps varies 
directly as the degree of solidity of the objects against which they 
happen to close their mandibles. 

Prosalience is exhibited by at least three very different groups 
of ants : certain bull-dog ants of the Australian genus Myrmecia, 
Gigantiops, and the extraordinary genus Harpegnathos, or Dre- 
panognathus, as it was formerly designated, of the Indomalayan 

The leaping Myrmecias, popularly known as " jumpers " in 
Australia, comprise the members of Emery's subgenus Pristo- 
myrmecia (fulvipes Rog., mandibularis Sm., and piliventris Sm.) 
and the smaller species of Myrmecia sens. str. allied to nigrocincta 
Sm. and pilosula Sm. I have frequently seen these ants jump 
distances varying from one to a few inches. When disturbed M. 
nigrocincta and pilosula, especially, present a ludicrous appearance 
as they bound out of their small mound nests- in a series of short 
hops like Lilliputian cavalry galloping to battle. Examination of 
these ants reveals a structural peculiarity which has .been over- 
looked by previous observers, namely, a distinct elongation and 
basal incrassation of their hind femora, as compared with the hind 
femora of the other nonsalient species of the genus. In Fig. 2 
the hind femur of a worker M. nigrocincta (c) and that of a small 
worker of M. sanguinea of the same size (d) are drawn to the 
same scale. The greater length and volume of the femur of the 
former species shows a distinct approach to the conditions in the 



saltatory Orthoptera and is evidently to be interpreted as an ar- 
rangement for the accommodation of a more voluminous and there- 
fore more efficient extensor, or abductor muscle in the hind leg. 
This morphological peculiarity, together with the leaping habit, 
seem to me to be sufficient to justify a separation of these jumping 
species from the remaining Myrmecias as a distinct subgenus, for 

FIG. 2. Femora in profile and cross section of five species of ants drawn 
to the same scale, a. Gigantiops destructor; b, Componotus castaneus amer- 
icanus Mayr ; c, Myrmecia nigrocincta Smith ; d, small Myrmecia sanguined 
Smith ; e, Harpegnathos saltator Jerdon. 

which the name Halmamyrmecia subgen. nov. (with pilosula 
F. Smith as subgenotype) may be proposed. 

In Gigantiops I find an even more pronounced elongation and 
basal incrassation of the hind femur than in Pristomyrmecia or 
Halmamyrmecia. As there is only one species of Gigantiops, I 
have compared its hind femur (Fig. 20) with that of a non-leaping 
Camponotus worker of the same size (b). It will be seen that 
the difference in the length of the two femora is very pronounced, 



but that the difference in the basal incrassation is not so striking 
when the parts are seen from the side. The shapes of the cross- 
sections of the bases of the two femora-, nevertheless, show that 
the Gigantiops femur is much more voluminous and therefore 
capable of furnishing attachment for a much larger and more 
powerful extensor muscle, as in the crickets and grasshoppers. 

FIG. 3. Female of Harpegnathos saltator Jerdon (after Mayr.). a, dorsal 
view ; b, lateral view ; c, head of same from above. 

Extraordinary feats of leaping are performed by the species of 
Harpegnathos, a genus confined to Indochina and the Philippines. 
In these ants (Fig. 3) the structure of the head is very singular, 
the eyes being very large, larger, in fact, than in any other 
Ponerine, and placed very far forward, and the mandibles are 
quite unlike those of any other known species. They are very 
long, separated at the base but approximated, arcuately curved 
upward, and gradually tapering toward their tips and finely serrate 
along their inner borders. Each is provided at the base with a 
large, flat, triangular tooth, projecting inward and somewhat 
downward and backward. In cabinet specimens the blades of the 
mandibles are applied to one another and the basal teeth overlap. 
The leaping habits of this insect have been observed by Lefevre, 


Lewis ('82), Wroughton ('92), and Bingham ('03). Lefevre 
states that it can make leaps of 20 to 25 cm., and Lewis saw it rise 
into the air to a height of 5 or 6 inches till exhausted, when its 
leap did not exceed an inch. Wroughton says : " The single speci- 
men of the genus, which I have had the luck to find, made leaps 
of a foot or 18 inches with perfect ease, exactly like a grasshopper. 
I had much trouble in securing this specimen, and, when I suc- 
ceeded, I found she could sting better than she could leap." Un- 
fortunately we have no observations on the modus saltandi of this 
curious insect. Its hind legs are really very short and even thinner 
than those of many nonsalient ants as indicated in the figure of the 
hind femur of a worker H. sdtator Jerdon (Fig. 2e}. As this 
ant is nearly of the same size as the specimens selected for the 
other femora (o-rf), and as the outline is drawn to the same scale,- 
it will be evident that no such feats of leaping as described by 
Wroughton can be performed with such appendages. We must 
conclude, therefore, that the mandibles are employed for this pur- 
pose, but how they function is a matter of pure conjecture till the 
living insect can be carefully studied. It is conceivable that when 
about to leap the Harpegnathos opens her mandibles slightly till 
the two basal teeth just barely touch, and that she presses the tips 
of the mandibles against the ground so that their long, slender, and 
probably very elastic blades are more arcuately bent. We may 
suppose, moreover, that if the two teeth are suddenly permitted to 
slide over each other, with a concomitant sudden unbending of the 
mandibles, the insect would be precipitated forward much like a 
very elastic strip of metal or whalebone bent in an arc and sud- 
denly released. It is less probable that the insect leaps by insert- 
ing the tips of the mandibles in the ground and bending them in 
the opposite direction, i.e., by more nearly straightening them and 
then suddenly allowing them to return to their original curvature. 1 

i Since this paragraph was written I find that Professor Forel (1921, p. 
47) gives some notes on the method of leaping employed by Harpegnathos. 
He states, apparently on the authority of some correspondent in India or 
China, that this ant " fait des bonds formidables de plus d'un metre a 1'aide 
de ses longues mandibules un peu recourbees en haut. La tete entiere se 
recourbe sous le corps, se rejettant ensuite en avant, un peu a la maniere du 
thorax de nos insectes d'Europe nommes taupins." 


In conclusion, two interesting matters of a more speculative 
nature may be briefly discussed : 

1. The various macrophthalmic, or large-eyed, tropical ants 
mentioned in this paper, namely, the Formicine genera Gigantiops, 
Opisthopsis, CEcophylla, Gesomyrmex, Dimorphomyrmex, Scmt- 
schiella and Myrmoteras, the Ponerine genera Myrmecia and 
Harpegnathos, and I may add also the whole subfamily Pseudo- 
myrminse, constitute only a small percentage of the more than 
10,000 extant species, subspecies, and varieties of Formicidse. In 
the great majority of forms the eyes and ocelli of the female, and 
especially of the workers, have undergone considerable reduction 
in size or have entirely disappeared. In the male, however, which 
is the more conservative sex in ants, the eyes and ocelli are always 
large. These facts, together with what is now known of the ants 
of the Baltic amber, suggest that not later than Cretaceous time 
the females and workers were also all large-eyed, like the Scoliidoid 
wasps from which the Formicidee are derived, and that they pre- 
served this condition till after the social habit and a wingless 
worker caste had been evolved. During the Eocene and owing to 
the further development of the peculiar nesting and foraging 
habits, the females and workers became increasingly microphthal- 
mic and anophthalmic till the small-eyed or blind condition became 
established in the majority of existing forms. 

2. It will have been noticed that all the known prosalient ants 
have very large, convex, and minutely faceted eyes, and that all 
belong to archaic genera. I have already discussed the antiquity 
of Gigantiops. The Myrmecias are, I believe, justly regarded as 
the most primitive of existing ants, survivors, without more than 
specific diversification, from the early Eocene or the late Creta- 
ceous. Harpegnathos includes only a few rare species of very 
restricted geographical range, evidently relicts on the verge of 
extinction. It would seem, therefore, that the leaping habit may 
have been much more general among the most ancient macroph- 
thalmic Formicidse, but had been abandoned as incompatible with 
a more highly developed social organization. This seems to be 
shown even within the genus Myrmecia, the larger and more domi- 
nant species of which no longer leap. That this habit should still 
persist, probably in a degenerate stage, in a few large-eyed forms, 


may be due to the fact that ants endowed with unusual visual 
powers can retain such a habit .with some impunity. Santschi 
('u) and Brun ('14) have shown that vision is an essential factor 
in the homing behavior of ants that do not adhere very strictly to 
the topochemical trail made by themselves or their fellows from 
and to the nest. The leaping habit, if preserved and assiduously 
practiced in small-eyed ants, would, of course, often render it 
difficult or impossible to find the nest by means of the antennal 
sense alone. 1 


Ashmead, W. H. 

'05 A Skeleton, of a New Arrangement of the Families, Subfamilies, Tribes 
and Genera of the Ants, or the Subfamily Formicoidea. Canad. Ent., 
1905, P'P- 381-384. 
Berlese, A. 

'20 Gli Insetti, Vol. 2. Milan, Soc. Edit. Libraria. 1920. 
Bingham, C. F. 

'03 Ants and Cuckoo Wasps. Hymenoptera, Vol. 2. In Blanford's Fauna 
of British India, including Ceylon and Burma. London, Taylor and 
Francis, 1903. 
Biro, L. 

'gya Pattago hangyak (Jumping Ants). Rovart. Lapok., 4, 1897, pp. 73, 74. 
*97b Biologische Mittheilungen aus New Guinea. Ill Springende Ameisen. 

Berliner Ent. Zeitschr., 42, 1897, pp. 136, 137. 
Borgmeier, T. 

'20 -Zur Lebensweise von Odontomachus affinis Guerin. Ihering-Festschrift. 

Zeitschr. Ver. Wiss. Kunst, 1920, pp. 31-38. 
Brun, R. 

'14 Die Raumorientierung der Ameisen und das Orientierungsproblem im 

allgemeinen. Jena, Gustav Fischer, 1914, viii + 284 pp., 51 figs. 
Cockerell, T;. D. A, 

'15 British Fossil Insects. Proc. U. S. Nat. Mus., 49, 1915, pp. 469-499, 

6 pis. 
'20 Fossil Artho'pods in the British Museum, I. Ann. Mag. Nat. Hist. (9), 

5, 1920, pp. 273-279, 3 figs. 
'21. Some Eocene Insects from Colorado and Wyoming. Proc. U. S. Nat. 

Mus., 59, 1921, pp. 29-3.9, i pl-. 9 text-figs. 
Donisthorpe, H. S. J. K. 

'20 British OligO;cene Ants. Ann. Mag. Nat. Hist. (9), 6, 1920, pp. 81-94, 
i pl. ' 

i The mollusks of the genus Pecten are interesting in this connection, 
since, as is well known, they are able to leap considerable distances through 
the water by rapidly opening and' closing the valves of the shell and have their 
mantle margins beset with large, highly specialized eyes. 


Emery, C. 

'91 I.e Formiche dell 'ambra siciliana nel museo mineralogico dell Univer- 
sita di Bologna. Mem. R. Accad. Sc. 1st. Bologna (5), i, 1891, pp. 
141-165, 3 pis. 
'93 Zirpende und springende Ameisen. Biol. Centralbl., 13, 1893, pp. 189, 


'97 Formicidarum Species Novae vel minus Cognitaa in Collectione Mussei 
nationalis Hungarici, quas in Nova Guinea, Colonla Germanica, col- 
legit L. Biro. Termeszetr. Fiizetek 20, 1897, pp. 571-599, 2 pis. 
'05 Deux Fourmis de 1'Ambre de la B'altique. Bull. Soc. Ent. France, 

1905, PP- 187-189, 2 figs. 

'91 Abh. Geol. Spezialk. Els. 3, 1891. 
Forel, A. 
"78 Etudes Myrmecologiques en 1878. Bull. Soc. Vaud Sc. Nat., 15, 1878, 

PP- 337-392, i pi. 
'93 Observations Nouvelles sur la Biologic de Quelques Fourmis. Bull. 

Soc. Vaud Sc. Nat. (3), 29, 1893, pp. 51-53. 
'930 Sur la Classification de la Famille des Formic'ides avec Remarques 

Synonymiques. Ann. Soc. Ent. Belg. 37, 1893, pp. 161167. 
'12 Formicides Neotropiques. Part VI. Mem. Soc. Ent. Belg. 20, 1912, 

pp. 59-62. 
'17 Cadre Synoptique Actuel de la faune universelle des Fourmis. Bull. 

Soc. Vaud Sc. Nat., 51, 1917, pp. 229-253. 

'21 Le Monde Social les Fourmis. Tome. I. Geneve, Kundig, 1921. 
Heer, 0. 

'49 Die Insektenfauna der Teriargebilde von Oeningen und von Radoboj 
in Croatien. II Neue Denkschr. Schweiz. Ges. Na-turw., n, 1849, pp. 
1-264, 17 pis. 

(Letter on Drepanognathus.) Ann. Soc. Ent. France (6), 3, p. 

Lewis, G. 

'82 (Letter on Histeridae, etc.). In Nouvelles et Faits Divers. L'.Abeille 

(4), 2, 1882, p. 155, 156. 
Mann, W. M. 

'16 The Ants of Brazil. Bull. Mus. Comp. Zool., 60, 1916, pp. 339-490 7 

Mayr, G. 

'67 Vorlaufige Studien fiber die Radoboj-Formiciden. Jahrb. K. K. Geol. 

Reichsan., 17, 1867, pp. 47-62, i pi. 
'68 Die Ameisen des baltischen Bernsteins. Beitr. zur Naturkunde Preus- 

sens. I. K. Phys. Oekon. Ges. Konigsb., 102 pp., 5 pis., 1868. 
'86 Notizen fiber die Formiciden-Sammlung des British Museum in Lon- 
don. Verh. zool. bot. Ges. Wien., 1886, pp. 353-365. 
'87 Siidamerikanische Formiciden. Verh. zool. bot. Ges. Wien., 37, 1887, 

pp. 5"-532. 

'93 Erganzende Bemerkungen zu E. Wasmann's Artikel fiber springende 
Ameisen. Wien. Ent. Zeitg., 12, 1893, p. 23. 


von Motschulsky, V. 

'55 Lettre a M. Menetries in Motschulsky's " Etudes Entomologiques," 

1855, pp. 8-38. 
"59 Briefliche Notiz iiber springende Ameisen. Stettin. Ent. Zeitg., 20, 

1859, p. 201. - 
Nietner, J. 

'58 Uber eine springende Ameise in Ceylon. Brief an Chr. Drewsen. Stet- 
tin. Ent. Zeitg., 19, 1858, pp. 445, 446. 
Roger, J. 

"63 Verzeichniss der Formiciden-Gattungen und Arten. Berlin, A. W. 

Schade, 1863. 
Santschi, F. 

'n Observations et remarques critiques sur le mecanisme de 1'orientation 

chez les fourmis. Rev. Suisse, 19, 1911, pp. 303-338, 6 figs. 
Smith, F. 

'58 Catalogue of Hymenopterous Insects in the Collection of the British 

Museum. VI. Formicidse, 1858, 216 pp., 14 pis. 
Wasmann, E. 
. '92 Einiges uber springende Ameisen. Wien. Ent. Zeitg., n, 1892, pp. 

Wheeler, W. M. 

'ob A Study of Some Texan Ponerinse BIOL. BULL., 2, 1900, pp. 131, 

10 figs. 
'14 The Ants of the Baltic Amber, Schr. Physik. *fekon. Ges. Konigsb., 

55, IPM, PP. I-I44, 66 figs- 
'17 The Australian Ant-Genus Myrmecorhynchtis (Ern. Andre) and its 

Position in the Subfamily Camponotina:. Trans. Roy. Soc. Austral., 

41, 1917, pp. 14-19, i Pi- 
'18 The Ants of the Genus Opisthopsis. Bull. Mus. Comp. Zool., 62, 1918, 

PP- 343-362, 3 pis. 
Wroughton, R. C. 

'92 Our Ants. Part I. Journ. Bombay Nat. Hist. Soc., 1892, pp. 13-60, 

2 pis.